Purpose: This study aimed to systematically evaluate the validity of variables related to pregnancy, delivery, and key characteristics of the infant in the Danish National Patient Register using maternal medical records as the reference standard.
Patients and methods: We reviewed medical records of 1264 women giving birth in the Region of Southern Denmark during 2017. We calculated positive (PPV) and negative (NPV) predictive values, sensitivity, and specificity to estimate the validity of 49 selected variables.
Results: The PPV was ≥0.90 on most pregnancy-related variables including parity, pre-gestational BMI, diabetes disorders, and previous cesarean section, while it was lower for hypertensive disorders, especially mild to moderate preeclampsia (0.49, 95% CI 0.32-0.66). Sensitivity ranged from 0.80 to 1.00 on all pregnancy-related variables, except hypertensive disorders (sensitivity 0.38-0.71, lowest for severe preeclampsia). On most delivery-related variables including obstetric surgical procedures (eg cesarean section and induction of labor), pharmacological pain-relief, and gestational age at delivery, PPV's ranged from 0.98 to 1.00 and the corresponding sensitivities from 0.87 to 1.00. Regarding infant-related variables, both the APGAR score registered five minutes after delivery and birthweight yielded a PPV of 1.00.
Conclusion: Obstetric coding in the Danish National Patient Register shows very high validity and completeness making it a valuable source for epidemiologic research.
Keywords: delivery; epidemiology; pregnancy; registries; sensitivity and specificity; validity.
© 2024 Herskind et al.